In my hospital, there is a general medical ward where rheumatology and endocrinology specialties work side by side and share beds. Medical patients are often complex, with various conditions requiring different specialists, so the patient mix is diverse. This means that even though I'm covering the rheumatology side, I may see patients admitted primarily with endocrine problems. We cross-refer to the respective specialists as needed.

Last week I worked with the rheumatology team in Deal ward. The consultant, Dr Williams, is one of the smartest and most academic consultants I've worked with. She's a tall, slim African woman who seems to have trained extensively in the UK, judging by her demeanour and accent. She's very sensible and caring towards her patients.

I learned a lot during my week of 12-hour shifts over 7 days. I'd like to share two cases that caught my attention and provide valuable reflective learning.

First case: When doctors don't know what's happening, the patient isn't improving, and nature seems to be taking its course. What would you do?

The patient was a woman approaching 60 who had been in hospital for several weeks with anasarca that was refractory to treatment. I took over her care mid-week until Sunday, my last shift. She had been on maximum diuresis with support from the renal team, experiencing fluctuating acute kidney injury but showing no improvement. Due to her history of alcohol consumption and alcoholic liver disease, she was initially treated for fluid overload with ascites secondary to chronic liver disease. However, imaging showed only anasarca without significant ascites. There was also no evidence of heart failure that could explain her condition. We explored renal and autoimmune causes of oedema, but no tests yielded results that could pinpoint a diagnosis. Nevertheless, she received high-dose furosemide, spironolactone, metolazone, and a trial of steroids—all without significant improvement.

On Saturday, when I was the only doctor on the ward, the family called to express their wish that we stop all interventions and focus on keeping her comfortable. Feeling uneasy with this request, I consulted a senior colleague who deferred the conversation to Monday when the primary team and consultant who knew the patient better would be available. Having cared for her over the previous few days, I understood what the patient and family wanted.

She was still relatively young, and we couldn't establish a diagnosis despite nearly two months of admission. At one point, I told my junior that I wished we had an excellent diagnostician like Dr House (from the TV series). When I first saw her at the beginning of the week, she was for full escalation of care. The next day, the family and patient agreed that CPR, intubation, and mechanical ventilation wouldn't be appropriate. By week's end, the family wanted to "pull the plug." This situation is challenging and makes the treating team feel like they've failed. Though mostly lethargic, she was clearly a lovely woman with sound judgment. She seemed concerned about being a burden to everyone, including her family, so she cooperated with caregivers and often agreed with their suggestions. During the week, she expressed a wish to die peacefully with her family present. The following day, feeling slightly better, she was content with continuing medical treatment, suggesting that stopping everything would be premature. This reinforces the importance of regularly reviewing patients' wishes as circumstances change.

Perhaps to ease my workload, she was transferred to the gastroenterology ward later in my shift, so I lost direct follow-up. A few days later, while working in the acute medical unit, a colleague from her former ward updated me. To my surprise, she had been admitted to the ITU, treated there for a few days, and died that day as we were checking on her progress. It's tragic, and I feel sorry for her and her family. I believe this wasn't how they wanted her final days or hours to unfold. Miracles do happen, and as doctors, we tend to fight against the "grim reaper" and hope our patients recover. Often we lose this battle, which is painful despite our best evidence-based efforts.

I discussed the case informally with a palliative care specialist nurse who suggested getting a clearer understanding of the patient's and family's expectations to better support them. I've observed that many consultants pursue maximum medical therapy to ensure they've done everything possible before transitioning to end-of-life care. This transition period can be unclear and difficult for junior and middle-grade NHS staff to navigate, particularly during weekends and holidays when staffing is minimal. We often engage in "parallel planning"—continuing active treatment while maintaining a low threshold for initiating end-of-life care if the patient deteriorates further. The challenge is that there's no clear boundary between life and death; some patients appear near death yet recover completely, while others go in the opposite direction. In most scenarios, the on-call medical team attending to a deteriorating patient faces the difficult decision of whether to withdraw treatment based solely on the patient's current critical condition. That's why we advocate for parent teams to explore patient and family wishes early in the medical journey and update the ReSPECT (Recommended Summary Plan for Emergency Care and Treatment) form for every patient at risk of deterioration and death. In theory, this form is intended for everyone under health and social care.

Case 2 involves an older woman admitted following a fall, with self-neglect and safeguarding concerns. At 69, it's debatable whether she should be considered elderly, but she appeared older and frail to me. During her first few days, she was quite unwell, confused, and unable to communicate properly. After we began heart failure treatment for a possible silent myocardial infarction, she became more alert and communicative. One morning, she caught my attention: Lady: "Can you please get the doctor for me?" Me: "I am your doctor for today. Tell me if there's anything I can help with." Lady: "Can you please tell my doctor to let me go and get on with my life? I'm ready to die."

Her words broke my heart. I immediately replied in a defensive tone, "You're not going to die. You will get better." The lady shook her head and said, "Let me get on with my life and give this bed to someone who really needs it." I felt emotional. This was only the second time in my two years working in the NHS that I'd heard a patient sensibly request to leave hospital and willingly give up their bed for someone else. Yet I've seen younger patients prolonging their hospital stays—sometimes for months—while waiting for government housing, care packages, or equipment. Some essentially live in the hospital while awaiting these support services. Since they're medically stable, they often leave the ward to smoke or disturb staff. Occasionally, they suffer negative consequences of extended hospitalisation, such as nosocomial infections requiring further treatment. I sometimes feel these patients are misusing the services we provide, preventing others from accessing beds and timely treatment. This represents a misuse of taxpayers' money.